Interestingly, the article explicitly mentions the FBI, not the NSA. But, if we were to unify them in our respective imaginations (for the purposes of discussion), they still need to have the budget for their operations supplied from somewhere. http://www.fbi.gov/news/testimony/fbi-budget-request-for-fis... - it appears as if the FBI's budget is just as rigidly controlled as any other governmental organisation. How they internally allocate the funding they receive is a different matter. To me it seems that if there is any form of oversight within the FBI, then the funding of a project of this cost/return ratio would be infeasible and therefore the project could possibly be shut down. There is the possibility that this is an exercise in propaganda and scare tactics: "We can read your emails, so don't try anything. Check out our power". However, because the internal mechanics of the FBI are necessarily not public knowledge this is all speculation and unsubstantiated opinion. But, the argument can be made that this very fact enables the FBI to get away with scare tactics. I still don't see an organisation with the real-world constraints of money being able to store the sheer volume of data that they claim to. For arguments sake, let's assume that the year-on-year rate of email generation remains uniform for the next decade. In one year, they need to have the ability to store X amount of emails. In a decade, they'll need 10X amount of storage. How are they going to store all that data?
"I don’t think they are filtering it. They are just storing it"
How much would it actually cost to store all the emails that American citizens send and receive? I find it difficult to visualise a system of checks and balances approving the massive budget required to house all that data. And the technical challenge of sifting through those emails would be seriously hard. I understand that this might not be just about national security, but it could also be a power game on the part of the FBI. One has to consider how much [national security | power] this really affords [US citizens | the FBI] when measured against the gargantuan expenditure required to actually pull it off. This makes me totally skeptical. Additionally, by saying 'basically the e-mails of virtually everybody in the country' Binney demonstrates his lack of conviction and uncertainty of his own claims.
So, if his words aren't the giveaway, then two minutes of critical thinking will make the interview seem alarmist and inaccurate.
Slightly OT:
"The Rods... so sensitive that they can be triggered by single photon"
This sort of thing is why I love HN. Irrespective of the topic, there is always some little gem I find somewhere. I was very sceptical of this claim, so I looked it up. Turns out it is possible:
So which framework comes out on top with the least developer friction metric? Not just convenient features like an ORM but also, your ability to do stuff without the framework getting in the way.
Are you sure? How do you know it is not an image processing/edge detection algorithm running? If they were indeed constructing a 3d-model, I'd be interested in seeing it, even if it is a low-poly approximation...
BTW, I've been meaning to ask for a while though: what sort of algorithm is running that detects edges and orientation so when the mouse is moved, the small white circle or quadrilateral seems to touch the objects in the image?
Initially this might seem different from the topic at-hand, but bear with me please:
Several years ago (probably first or second year of high school) I wrote and distributed amongst my friends a small desktop game. The program e-mailed me their new high-scores and internal game diagnostics daily. I found that this was too often so I rewrote the game to send these data daily. This way of retrieving data sucked. Often, I would get empty or unimportant messages.
Eventually I had the realisation that time-oriented polling was the wrong philosophy. Instead, there should be a "this event has happened now" (e.g. 500kb of diag. generated) algorithm that calls the polling routine/invokes the 'stuff to be done traverser'. If I understand the problem correctly, the mass e-mail example, you could make a simple counter as part of the HTTP request. When this counter hits a certain you-determined threshold value (e.g. there are 500 jobs to do), then call the traverse/processing code:
if jobs_to_do > thresh then invoke some async processing
The takeaway from this post is that, in my opinion, time-based (cron job-bish) polling algorithms are inefficient but can be replaced with a rudimentary event-driven ones.
And just BTW, 'countervailing' - I'm deifying this word. Holy s*. I'm going to use it all the time.
A comparison between this and statsd would be interesting. Any one have experience with both?
And as for the similarity between the Heroku documentation page - their github documentation page reads:
"This program is forked from heroku/heroku-docs, and originally written by @rtomayko and @adamwiggins. Later, modified by @kzk and @doryokujin."
I think this article's headline is a little misplaced. The massive backing given to Ouya on Kickstarter in such a short time period is testament to the sheer volume of people who want indie gaming hardware. Gamers aren't going to throw around terms like 'saturated market' and 'product differentiation' when a $99 game console comes out.
Don't forget RethinkDB, Mixpanel and Cloudera - those guys are solving some difficult, real world problems. Personally, I wouldn't classify massively distributed computation as a service boring. If anything, the scale (in numbers of users, amount of data) makes B2B of this kind even sexier than share-my-cat-at-my-location-agram.
Kudos to this guy for giving a head start to plenty of indie and starting game devs. I can attest to this as I used Cocos2D to produce my iOS game (since pulled because I have been too occupied this year to hack on iOS: last year of high school = major pressure to do well; very competitive environment). But this is beside the point: Wenderlich's tutorials are awesome, well-written and comprehensive. They won't get you production ready games in a flash, but they sure do help.
From an Indie point of view, iOS game development is considerably more accessible than Android game development by virtue of the following facts:
* Cocos2D is a robust, free and easy-to-use environment; a fully-functional, stable, Android analogue does not exist.
* iOS code is 99% universal (just works on all devices) whereas Android code really isn't. For instance: OpenGL sample code from the Android development guides (http://developer.android.com/training/graphics/opengl/index....) does not work on 2 of 4 android devices I've tried it on because it targets deprecated API's and methods which are unusable using the newest SDK versions (it was version 14 or 15 at the time; largely unimportant). Imagine how disheartening, demotivating and demoralising it is to see that "official Google endorsed code" just doesn't work.
* I have personally tried AndEngine and Cocos2D for Android. Neither is up to scratch with iOS Cocos2D.
tl;dr Ray Wenderlich is indirectly responsible for a ton of Indie games on the App Store, should be thanked, and no, an equal (easy-to-use portable, robust, _decent_ framework) doesn't exist for android (though there may be analogues).
The $99 is not so much the limiting reagent... South Africa doesn't mean poor BTW :) its just the barrier to entry that is a tad high for an income-less High School student.